Mg2+ is an essential element. The basis of Mg2+ selectivity, binding and coordination with enzymes and small molecules is well understood but the molecular basis by which transport systems achieve Mg2+ selectivity over other cations is not understood in any system. The structures of the CorA and MgtE Mg2+ channels demonstrate clearly that channel selectivity for Mg2+ cannot be explained by analogy to the chemistry of selective Mg2+ sites in any enzyme system. The CorA Mg2+ channel is the primary source of Mg2+ for ~50% of all Bacteria and Archaea while MgtE is the primary source for the other half of each kingdom. Both have multiple homologs in eukaryotes, including humans. Electrophysiological data show that both CorA and MgtE are true ion channel with surprisingly high conductance, >100 pS, a flux rate >107 Mg2+ ions/sec. This very high rate poses serious mechanistic issues. The rate of dehydration of Mg2+ is only 105/sec, 100-1000-fold slower than the flux rate. Uniquely among all known ion channels, both CorA and MgtE initially bind a fully hydrated Mg2+ ion, which must contribute to the channel's selectivity for Mg2+. To mediate flux of a largely dehydrated cation, the channel must in some manner accelerate the rate of dehydration. However, electrostatic interactions are not involved in this process or in ion flux. The study of these two ion channels, which have quite different structures, provides an opportunity to dissect the distinct chemistry involved in Mg2+ selectivity. Aim 1 will investigate the mechanism of selectivity of CorA for Mg2+ through i) electrophysiological analysis of site-directed mutants expressed in giant cells or reconstituted in lipid, ii) hydrogen-deuterium exchange primarily of loop and pore residues, and iii) X-ray crystallographic approaches using cations with very slow rates of dehydration and cysteine mutations in transmembrane segment 2 of CorA. The latter generate crosslinked oligomers which appear to be trapped in a conformation different from the closed state, previously solved. Their crystallization may provide additional conformational information for CorA. Aim 2 will explore analogous issues for MgtE using the same approaches.